Azure OpenAI Service
An Azure service that provides access to OpenAI’s GPT-3 models with enterprise capabilities.
3,783 questions
Fine-tuning Availability for OpenAI Models in Azure
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(108.80000000000001, 74%, 20%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EAF%3C/text%3E%3C/svg%3E)
Alison Felix
0
Reputation points
Is it currently possible to fine-tune OpenAI models like GPT-4, GPT-4o, and GPT-4o-mini? I have been attempting to fine-tune these models over the past few days in the North Central US and Sweden Central regions, which are listed as available for fine-tuning here: Supported Models for Fine-tuning.
However, I consistently encounter the following error when uploading the training data:
I have attempted the role assignment mentioned in the error message without success, and the file format is correct, as I can successfully upload it for fine-tuning other models, such as Phi-3-mini-4k-instruct. Additionally, I tried using Machine Learning Studio, but the GPT models also appear to be unavailable for fine-tuning in both regions.
What are the next steps to resolve this issue, or are the OpenAI models temporarily unavailable for fine-tuning?
{count} votes
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(166.4, 22%, 26%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ESC%3C/text%3E%3C/svg%3E)
SriLakshmi C 3,245 Reputation points Microsoft External Staff2025-03-05T07:34:49.17+00:00 Hello Alison Felix,
Greetings and Welcome to Microsoft Q&A!
I understand that you're encountering issues with fine-tuning OpenAI models like GPT-4, GPT-4o, and GPT-4o-mini in the North Central US and Sweden Central regions. Please consider these steps:
Ensure that your role assignment, data quality, and API/SDK versions are correctly configured. Verify that your account has the necessary permissions, such as the Cognitive Services OpenAI Contributor or Owner role, assigned at the appropriate scope (subscription, resource group, or specific resource).
Confirm that your training dataset meets the required JSONL format, is properly structured, and does not contain malformed entries, as data quality issues can cause upload failures.
Check that you are using the latest version of the Azure OpenAI API and SDK, as outdated versions may lack support for fine-tuning or contain unresolved bugs.
If the issue persists, consider testing in another supported region.Kindly refer this Troubleshooting for Azure OpenAI fine-tuning.
I hope you understand. And, if you have any further query do let us know.
Thank you!
-
SriLakshmi C 3,245 Reputation points Microsoft External Staff
2025-03-06T09:40:00.3566667+00:00 Hi Alison Felix,
Did you get any chance to check the above response. Thank you!
-
SriLakshmi C 3,245 Reputation points Microsoft External Staff
2025-03-07T09:07:36.26+00:00 Hi Alison Felix,
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Thank you!
-
Alison Felix 0 Reputation points
2025-03-07T11:16:28.2933333+00:00 Still have the same issue. I've tried it again today both in North Central US and Sweden Central with all the suggested role assignments but I still can't fine-tune GPT-models. The training dataset still works for other models so I don't think it's causing the issue.
-
Alison Felix 0 Reputation points
2025-03-07T11:19:21.05+00:00 Hello @SriLakshmi C ,
I tried it again today but still have the same issue. This time I've also added the roles Cognitive Services OpenAI Contributor and Cognitive Services Contributor to my user at the resource group level and at the AI Services level, as suggested in the Troubleshooting link you provided, still with no success. The file I'm using has worked in the past when fine-tuning GPT models and still works when uploaded to fine-tune other non-GPT models. I've also tried both North Central US and Sweden Central again.
-
SriLakshmi C 3,245 Reputation points Microsoft External Staff
2025-03-12T07:57:38+00:00 Hi Alison Felix,
Apologize for the delayed response,
To resolve the fine-tuning issue,
Verify the availability of the specific OpenAI models in your selected region and confirm that fine-tuning is supported for those models.
Check for any service limitations or quota restrictions that might be preventing the fine-tuning process.
Ensure that your subscription has the necessary capacity for fine-tuning requests. If the issue persists, consider exploring alternative regions where fine-tuning is available and test the process in a different supported location to determine if the problem is region-specific.
Thank you!
Sign in to comment
Sign in to answer